How do you calculate LTV for a Recurring service that accounts for outliers?

We’ve been testing a Subscription version of our software for a few years now.

I went in and calculated what I think our LTV per customer is.
I looked at a group of customers who subscribed in 2013 and then looked at their lifetime value …so far.
So I calculated total months, total revenue and then divided by number of customers.

BUT… here’s the kicker:
About 6% of those users are still subscribed. They could stay subscribed forever.
I don’t have enough data to predict that (based on the math I know :slight_smile:

Any suggestions?

Calculate monthly churn and extrapolate from there.

1 Like

Why worry about the outliers? Use your calculated LTV number as a “good enough” number - you’ll be erring on the side of being conservative with it, which will prevent overspending on acquiring customers, etc.

True, but I run the risk of missing marketing opportunities. (As you have suggested, I am usually very conservative, but that can lead to missing opportunities)

Let’s say my conservative LTV is $100 but the true value is $200. And I find that Adwords costs $110 per paying customer.
If I use the conservative number, I’ll never advertise. But I might miss out on nearly doubling every dollar of advertising I put into it.

And… I was calculating LTV wrong. @pjc correctly pointed out that I need to base it on Churn rate.